Roboflow MCP Server with 29 Tools for Claude, Cursor, and AI Agents
Manage computer vision workflows — upload images, train models, and manage datasets directly from your AI agent. Vinkius routes your AI agents directly to Roboflow through a governed connection. 29 tools ready to use with Claude, ChatGPT, Cursor, or any AI agent — no hosting, no setup, connect in 30 seconds.
Ask AI about this server
Compatible with every major AI agent and IDE

* Every MCP server runs on Vinkius-managed infrastructure inside AWS - a purpose-built runtime with per-request V8 isolates, Ed25519 signed audit chains, and sub-40ms cold starts optimized for native MCP execution. See our infrastructure
What is the Roboflow MCP Server?
The Roboflow MCP Server routes AI agents like Claude, ChatGPT, and Cursor directly to Roboflow via 29 tools. Manage computer vision workflows — upload images, train models, and manage datasets directly from your AI agent. Powered by Vinkius — your credentials stay on your side of the connection, every request is auditable. Connect in under 2 minutes.
Built-in capabilities (29)
Tools for your AI Agents to operate Roboflow
Ask your AI agent "List all projects in my Roboflow workspace 'industrial-safety'." and get the answer without opening a single dashboard. With 29 tools connected to real Roboflow data, your agents reason over live information, cross-reference it with other MCP servers, and deliver insights you would spend hours assembling manually.
Works with Claude, ChatGPT, Cursor, and any MCP-compatible client. Powered by Vinkius — your credentials never touch the AI model, every request is auditable. Connect in under two minutes.
Why teams choose Vinkius
One subscription gives you the infrastructure to connect your AI agents to thousands of MCP servers — and deploy your own to the Vinkius Edge. Your credentials stay yours. Your data flows directly between your agent and the API. DLP blocks sensitive information from ever reaching the model, kill switch for instant shutdown, and up to 60% token savings. Enterprise-grade routing and governance, zero maintenance.
Build your own MCP Server with our secure development framework →The Roboflow App Connector works with every AI agent you already use
…and any MCP-compatible client


















Use all 29 Roboflow tools with your AI agents right now
Vinkius routes your AI agents to Roboflow through a governed proxy. Beyond a simple connection, you get full visibility into every action your agents perform, with enterprise-grade security and up to 60% savings on AI costs.
Add projects to folder on Roboflow
Add projects to a folder (Enterprise)
Auto label on Roboflow
Start an auto-labeling job using foundation models
Cancel training on Roboflow
Cancel an active training job
Create annotation job on Roboflow
Assign a batch of images to a labeler and reviewer
Create folder on Roboflow
Create a project folder (Enterprise)
Create project on Roboflow
Create a new project in a workspace
Delete images on Roboflow
Delete multiple images from a project
Delete project on Roboflow
Delete a project or version (moves to Trash)
Download dataset on Roboflow
Retrieve a download link for a zipped dataset in a specific format
Fork universe project on Roboflow
Fork a public project from Roboflow Universe
Get async task on Roboflow
Track long-running operations like forking or large exports
Get dataset health on Roboflow
Check dataset health (class distribution, missing annotations, etc)
Get image on Roboflow
Get details for a specific image
Get project on Roboflow
Get project details, metadata, and versions
Get root on Roboflow
Verify authentication and retrieve default workspace
Get training results on Roboflow
Retrieve metrics and status for a version training run
Get version on Roboflow
Retrieve metadata for a specific dataset version
List folders on Roboflow
List project folders in a workspace (Enterprise)
List trash on Roboflow
List items in the workspace trash
List workspace projects on Roboflow
List information about a workspace and its projects
Manage image tags on Roboflow
Add, remove, or set tags on an image
Restore trash on Roboflow
Restore an item from the trash
Run inference on Roboflow
Run inference on an image using hosted models
Search project images on Roboflow
Search and filter images within a specific project
Search workspace images on Roboflow
Search and filter images within a workspace
Start training on Roboflow
Start training a model on a dataset version
Stop training on Roboflow
Early stop an active training job
Upload annotation on Roboflow
Attach an annotation file to an existing image
Upload image on Roboflow
Upload an image to a project
What the Roboflow MCP Server unlocks
Connect Roboflow to your AI agent to streamline your computer vision pipeline. From dataset management to model training and inference, handle your entire CV lifecycle through natural language.
What you can do
- Workspace & Project Management — List projects, create new ones, or fork from Roboflow Universe to jumpstart your development.
- Dataset Operations — Upload images (via URL or Base64), manage versions, and download datasets in various formats like COCO or YOLO.
- Model Training — Start training runs, monitor results, and retrieve precise performance metrics (mAP, precision, recall) for any version.
- Image Search — Search and filter images within your workspace to audit your data and improve model accuracy.
- Inference & Results — Run inference on images and retrieve results to verify model behavior in real-time.
How it works
1. Subscribe to this server
2. Enter your Roboflow Private API Key
3. Start building and managing vision models from Claude, Cursor, or any MCP-compatible client
Who is this for?
- ML Engineers — monitor training progress and dataset health without leaving the terminal or IDE.
- Data Scientists — quickly query dataset versions and export data for custom training scripts.
- Product Teams — audit model performance and visualize inference results through simple conversation.
Frequently asked questions about the Roboflow MCP Server
How can I verify if my Roboflow API key is correctly configured?
You can use the get_root tool. It will attempt to authenticate with your key and return the default workspace details if successful.
Can I get the training performance metrics for a specific model version?
Yes! Use the get_training_results tool by providing the workspace, project, and version ID. It returns mAP, precision, recall, and other training metrics.
Is it possible to export my dataset to a specific format like YOLOv5?
Absolutely. Use the download_dataset tool and specify the format parameter (e.g., 'yolov5pytorch') to receive a download link for your zipped dataset.
More in this category

Paperspace
6 toolsProvision and track powerful GPU workloads via Paperspace — list compute instances, fetch active deployments, trace team projects, and query Gradient environments via AI.

Transloadit
12 toolsVersatile file uploading and processing API.

CockroachDB Cloud
8 toolsManage distributed SQL clusters via CockroachDB Cloud — track clusters, monitor nodes, and audit network allowlists directly from any AI agent.

Webflow Alternative
19 toolsManage Webflow sites, CMS collections, pages, and assets directly from your AI agent.
You might also like

Flux Markets
9 toolsAccess energy market data, live pricing, and proprietary trade insights from Flux Markets.

Pinterest Ads
8 toolsEquip your AI agent with direct access to Pinterest Ads — manage campaigns, track pin performance, and optimize shopping ad spend without opening Pinterest Ads Manager.

Missing Value Imputer
1 toolsAutomatically fill NaN and missing values in datasets using Mean, Median, Mode, or Zero strategies deterministically local. Essential ML data preparation.

Apex & Division Stats by TRN
9 toolsDeep gaming analytics — player profiles, per-legend breakdowns, match history, and global leaderboards for Apex Legends and Division 2 via AI.
We built the connector to Roboflow. Now put your agents to work. Fully governed.
Vinkius is the AI Gateway with managed hosting. Stop building connectors. Every connection runs inside eight layers of security.
Hosted, sandboxed, and live on AWS. You don't provision anything. You don't maintain anything. You connect.
Every tool call, every token, every response. Logged and auditable. Data flows direct from Roboflow to your agent. Nothing is stored on our side. Ever.
Eight governance layers on every request. Sensitive data redacted before it reaches the model. Kill switch if anything goes sideways. Always on.
